This study is a double-blind, randomized, clinical trial to determine the effectiveness of intermittent use of oral diazepam (Valium) in preventing recurrences of febrile seizures. Preventing recurrences is desirable: seizures are upsetting to families, result in medical costs, may cause physical injury, and might result in brain damage. In addition, there is a relationship between frequency of febrile seizures and later development of epilepsy. Children six months to five years old, with a history of at least one febrile seizure, no previous afebrile seizures, and not on anticonvulsants, will receive either diazepam or an identically-appearing placebo. Following ascertainment of eligibility and informed consent, children will undergo a baseline medical history, physical and neurologic examinations. Diazepam or placebo will be given to the child during subsequent febrile episodes. Drug compliance will be monitored by urine testing. Patients will be followed carefully with regular telephone calls to parents to ascertain the occurrence of fevers and/or seizures. Patients who have additional seizures will undergo further neurologic examination. Patients will be recruited from four study sources: a pediatric primary care clinic, a pediatric group practice, a pediatric inpatient service and a pediatric neurologist's office practice. There will be a central data coordinating center for processing and statistical analysis of information collected. One hundred sixty children will be enrolled over a two-year period. Each subject will have a minimum of one-year's follow-up for recurrence of febrile seizures. Statistical power considerations, based on an anticipated one-year recurrence rate of 30-50% in the placebo-treated group, indicate that the sample size is sufficient to detect with high probability (power .80 or higher) a true diazepam treatment effect that halves the one-year recurrence rate for febrile seizures. The results of this study should have important implications for the prevention of febrile seizure recurrences and accompanying psychologic distress, medical costs, and possible neurologic sequelae.